LONDON – The EMA’s expedited review of potential Ebola virus infection treatments has concluded there is not enough data as yet to support the safety or efficacy of any of the seven products it has assessed.

The news about these non-approved products coincided with the publication of a paper by U.S. and Canadian researchers who have found 53 marketed drugs showing activity in an in vitro cell model of Ebola infection that they say could be suitable for repurposing.

Commenting on the EMA review, the agency’s head of anti-infectives and vaccines, Marco Cavaleri, made a plea for companies to come up with more data, saying, “We encourage developers to generate more information on the use of these medicines in treating Ebola patients.” The EMA will review new data as soon as it becomes available, Cavaleri said.

The seven products in question are Biocryst Pharmaceutical Inc.’s BCX 4430, a broad-spectrum antiviral that inhibits RNA polymerase; brincidofovir from Chimerix Inc., which is also a polymerase inhibitor; Mapp Biopharmaceutical Inc.’s monoclonal antibody Zmapp; Sarepta Therapeutics Inc.’s viral RNA inhibitor AVI-7537; Toyama Chemical and Medivector Inc.’s, broad-spectrum antiviral favipiravir; TKM-100802, an siRNA drug being developed by Tekmira Inc., of Vancouver, British Columbia; and an anti-Ebola polyclonal antibody Fab, proposed by Fab’entech, of Lyon, France.

The file of data behind each of the seven products is highly variable. While some have not been tested in humans, others have been administered to Ebola patients on a compassionate basis or have been in human trials for the treatment of other diseases.

Of the products that have been used on a compassionate basis, the EMA said the available clinical data are not sufficient for an evaluation of efficacy, safety, and pharmacokinetics in the target population, at this stage. The appropriate dosing of all the products reviewed is uncertain.

These may not seem surprising conclusions, given that all the products are at various stages of clinical development (apart from Fab’entech’s equine antibody Fab, which is a proposed product based on an existing platform technology).

The EMA began the review in September after the World Health Organization ruled the current epidemic means it is ethical to test non-approved products in patients, and began laying the ground for trials. The review is intended to provide independent information for health authorities faced with the decision of whether or not to allow testing of an experimental treatment.

A spokeswoman for the EMA stressed the aim is to provide independent advice, and told BioWorld Today, “The review does not aim to make recommendations on clinical trials of specific medicines.”

The review includes an assessment of the validity of animal models of Ebola virus, which could be useful to anyone interested in repurposing any of the 53 marketed drugs that U.S. and Canadian researchers have identified as blocking Ebola virus-like particle entry into cells in a high throughput preclinical screen.

The 53 products identified in the screen include microtubule inhibitors, estrogen receptor modulators, antihistamines, antipsychotics and ion channel antagonists. Several of the drugs were already known to inhibit entry of Ebola virus particles into cells, indicating the miniaturized screen is a useful way to identify drugs for repurposing.

Details of the hits are published in the Nature journal, Emerging Microbes and Infections. The researchers say they have decided to publish all the data to allow other scientists to test the compounds in wild-type Ebola virus infection assays and animal models, to confirm the preliminary findings.

While the EMA cites uncertainties about dosing as an issue for all the seven experimental products in its review, other shortcomings vary, and supply is an issue for most.

Biocryst’s BCX4430 has shown an effect on survival in rodent models. Studies are ongoing in Ebola challenge models in non-human primates, however 17 out of 18 non-human primates given BCX4430 survived in a Marburg virus trial. Supplies of the drug are limited.

Brincidofovir meanwhile has only in vitro efficacy data in Ebola. However, the product is effective in treating other viruses in rodent models. Over 900 subjects have received brincidofovir, including a few patients with Ebola virus infection, but no safety data was available for review and the EMA says it does not have enough information to establish a suitable dose for Ebola infection.

Favipiravir is approved in Japan for treating influenza and has approvals for phase III studies in Europe. To date 2,500 people have received the drug, but there is no Ebola data in non-human primates or humans, even though a few Ebola patients have been treated with favipiravir. It is suggested the dose needed to treat Ebola will be higher than flu, but the EMA said there is not enough information to decide on dose.

TKM-100802 is not optimized for the Guinea strain of the virus that is causing the current epidemic. In three non-human primate studies against other strains, TKM-100802 was effective when administered up to three days post-infection.

The FDA placed the product under a clinical hold in July after one healthy volunteer who received the highest dose in a 14-subject phase I trial suffered cytokine release syndrome. In August the FDA changed this to a partial hold to allow TKM-100802 to be used to treat Ebola patients. The EMA said it has not seen any safety data relating to this compassionate use.

While AVI-7537 enhanced survival in non-human primates models of Ebola infection, most of the data were generated with a previous compound, AVI-6002, a 1:1 mix of AVI-7537 and AVI-7539. (AVI-7539 was later shown not to be active.) No patients have been treated with the product.

The data file for Zmapp includes information on seven patients treated with the product at different stages post-infection, of whom five survived. In non-human primate trials all the 18 animals treated with Zmapp survived. However, the EMA said “no conclusion of its suitability for human use can be made” at this stage.

Although non-human primates are seen as good models for human infection, the EMA said these challenge studies need to show a product is effective once the virus has multiplied in the host and is producing symptoms, rather than immediately after infection. Only Zmapp and TKM-100802 have been shown to have an effect when administered at some time after the viral challenge.